The psychometric reliability and discriminant validity of a computerized, self-report personality interview was assessed. The subject sample (N=100) consisted of male and female students from a university introductory psychology course, employees from a general hospital and volunteers from a metropolitan community. Subjects completed the Computerized Personality Disorder Interview (C-PDI) on two occasions separated by an interval of one week. Diagnoses were compared with scale score means. Kappa coefficients for diagnosis ranged from .10 (Avoidant) to .78 (Schizoid). Test-retest correlation coefficients for scale score means ranged from .74 (Obsessive Compulsive) to .90 (Schizotypal and Borderline). Males were significantly higher than females for the Schizoid and Narcissistic scales (p<.01). The C-PDI discriminated between subjects with psychiatric, substance use/abuse, and family psychiatric histories. The advantages of a dimensional system of classification for personality disorders which uses scale scores are discussed.